Abstract
To the Editor:
Case Report
J. is a 13-year- old African American boy with attention-deficit/hyperactivity disorder (ADHD) combined type, oppositional defiant disorder, mixed receptive-expressive language disorder, learning disorder not otherwise specified, and insulin-dependent diabetes mellitus (onset 2 years previously) admitted to an inpatient psychiatric unit with worsening aggression at home and school. J. had a longstanding history of emotional and behavioral dysregulation, especially in response to parental limit setting, with low frustration tolerance, difficulty with social cognition and communication, and ADHD symptomology (inattention, hyperactivity, and deficits in organization and cognitive shifting). He had a history of failed psychostimulant and behavioral intervention trials.
In the weeks prior to inpatient psychiatric hospitalization, J. demonstrated worsening agitation and oppositional behavior, characterized by verbal aggression, running away from home, property destruction, and physical aggression toward others. Psychiatric hospitalization occurred in response to a behavioral outburst during an outpatient therapy session, which resulted in parental injury and physical restraint of the patient. Prior to hospitalization, J. was prescribed no psychotropic medications and had no history of antipsychotic medication trials.
At psychiatric admission, J. was a healthy-appearing teenage boy with a weight, height, body mass index, and hemoglobin A1C of 50.8 kg, 151 cm, 22.3 kg/m2, and 7.5%, respectively. The patient was transitioned from his home insulin pump to basal-bolus insulin via syringe, with insulin lispro given before meals and insulin glargine given at bedtime. He was placed on a diet for people with diabetes. Glycemic monitoring and insulin correction and administration occurred only through nursing staff.
By day 2 of hospitalization, J. demonstrated good glycemic control, with measured blood glucose levels largely under 100 mg/dL. On hospital day 3, he was placed on standing risperidone 0.5 mg at bedtime. This dose was maintained for 3 consecutive days. With no changes in diet and activity or reductions in insulin, blood glucose measurements were noted to be consistently elevated during this short risperidone trial, with blood glucose levels often >200mg/dL. After risperidone discontinuation, blood glucose levels decreased and returned to baseline ∼36 hours after the last antipsychotic dose.
Aripiprazole 2.5 mg each morning was initiated 2 days after risperidone discontinuation. Prior to aripiprazole administration, J. had demonstrated good glycemic control for over 24 hours. After aripiprazole initiation, serum glucose measurements showed slow elevation to readings consistently >200 mg/dL by <36 hours after the first dose. This aripiprazole dose was maintained for 6 days. Following aripiprazole discontinuation, blood glucose levels again showed decline. Other medications initiated during J.'s hospitalization were guanfacine 0.5 mg twice daily, ergocalciferol 1000 U daily, and diphenhydramine 25 mg at bedtime for sleep, all of which were well tolerated and consistently dosed. There was no temporal relationship between glycemic control and administration of these medications noted. The patient's discharge weight was 52 kg.
Discussion
Although the adverse metabolic effects of atypical antipsychotics are not fully understood, there appear to be both short- and long-term metabolic sequelae. One study found that 15% of subjects who were antipsychotic-naïve developed impaired fasting glucose in their initial period of antipsychotic titration (Ryan et al. 2003). After being matched for lifestyle factors and body habitus, recipients of antipsychotics in this study were found to have significantly higher cortisol levels than control subjects. Notably, this finding is not consistent with what is known about long-term chronic antipsychotic use, where hypercortisolemia is not typically seen (Newcomer and Haupt 2006). These observations suggest that antipsychotic medications cause a transient rise in cortisol, which may, in part, account for initiation-related hyperglycemia. The longer term metabolic syndrome may be unrelated to elevations in cortisol.
There is a well-known relationship between adiposity and dyslipidemia (which are known to occur as a result of atypical antipsychotic administration) (Newcomer et al. 2002) and insulin resistance (Reaven 1999). One could speculate that alterations of lipid levels in the blood, which could occur immediately after antipsychotic treatment, might contribute to insulin resistance before the classic metabolic syndrome signs, such as weight gain, become apparent. There is evidence for this decoupling in an animal model study that showed a dose-dependent reduction in insulin sensitivity within 2 hours of initial exposure to clozapine or olanzapine (Houseknecht et al. 2005). Additionally, Jin et al (2002) found that ∼50% of new-onset diabetes cases in this population showed any significant weight gain at the time of initial hyperglycemia. This concept would be consistent with the finding that even exogenous insulin administration, as described in the case just described, may not function fully well in transporting glucose out of the blood stream and into cells. One potential explanation surrounds the ability of atypical antipsychotics to suppress the 5-HT2A receptor-mediated glucose uptake in muscle or adipose tissue, which is an immediate effect of medication administration (Gilles et al. 2005). Bucholz et al (2008) have speculated about another possible mechanism for immediate insulin resistance with atypical antipsychotics involving either direct inhibition of hepatocyte nuclear factor-1 α or inhibition of this transcription factor involved in glucose regulation through induction of cytochrome P450 1A2 (CYP1A2). Whether a multifactorial combination of these speculated factors or other unknown aspects are involved, it is evident that atypical antipsychotics can have an immediate effect on blood glucose levels, and these potentially deleterious outcomes must be considered in high-risk patients.
Footnotes
Disclosures
Drs. Stern and Trieu have no conflicts of interest or financial ties to report.